



















































ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

13

Community of Inquiry as a Complex 
Communicative System1

Nadia Stoyanova Kennedy

 

Complexity theory has been steadily gaining recognition over the last half a century as a new way of under-
standing the organization of systems on all levels—whether organisms, social systems, business organiza-

tions, or ecosystems. Contrary to a mechanistic, analytic approach to the study of phenomena, the complexity 
perspective takes into account all interactions between the components of a system, which it understands as in 
dynamic interplay. In this paper, I will attempt to re-describe community of inquiry (CI) as a complex system, 
and draw some implications that follow from taking into account such description.

By community of inquiry, I mean a relatively small (10-20) group of people engaged in a conversation (or, 
preferably, an ongoing series of conversations) about an agreed-upon concept or problem or question, which is 
convened and overseen by a “facilitator” who is committed to certain normative ideals, chief among which is the 
construction of an “ideal speech situation” (Habermas, 1990)—that is, a discursive setting in which everyone has 
equal right and opportunity to speak, in which intimidation of any sort is absent, and in which epistemological 
authority is distributed rather than centralized in one person. The group enters the dialogue with the shared 
intention of participating in ongoing critical deliberation together about the issues, with the expectation that 
new meaning will arise from their interaction and the challenging and testing of each other’s assumptions in the 
common space of dialogue. In all forms of community of inquiry there is a set of basic discourse rules, implicit 
or otherwise, that participants abide by, and one or more “facilitators”--who act to clarify and to coordinate the 
emergent structure of ideas and arguments that the conversation generates (Kennedy& Kennedy, 2010).

I will start by spelling out a distinction made by Paul Cilliers between two types of systems: complex and 
complicated (1998).  A modern jumbo jet is a complicated system—it has intricate sets of mechanisms and auto-
mated devices. Because all of those mechanisms have fixed relationships between their elements, the entire jet 
can be analyzed by “parsing” it into its different segments, and a thorough description and understanding of the 
jet can be produced by assembling the descriptions of the different parts that comprise it. On the other hand, 
an ecosystem or the brain are complex systems, in that they are characterized by non-linear relationships and 
feedback loops. Such systems can only yield partial analysis, which can produce partial descriptions. It is not pos-
sible to fully understand a complex system, like a cell, an ant colony, human society, or communication among a 
group of people by taking it apart like a mechanistic device. The traditional analytic method that works well with 
complicated and simple systems does not fare well with complex systems. The problem is that the traditional 
analytic approach of taking and studying smaller parts of a system destroys its so-called “emergent properties”, 
which result from the interaction among its components. As we already mentioned, those properties cannot be 
reduced to the characteristics of the components themselves. An example of an emergent property is conscious-
ness, something that arises from the interactions of single neurons and even mundane phenomena, such as 
walking, and which cannot be understood when only parts of the human body are examined. 

In community of inquiry, some emergent products are collective arguments or a collectively elaborated con-
cept, and even more ephemeral products like justice, respect, care.  Emergent properties are usually the impor-
tant features of the system that we want to study, and in order to do this, we have to deal with the system as a 
whole. As the truism associated with complex system goes, “the system is always more than the sum of its parts.” 



ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

14

Complex systems are typically the ones associated with living organisms or groups of living things like humans, 
viruses, social systems, language, oceans, or the brain--systems, that shift and change, self-organize and self-pro-
duce, of which community of inquiry (CI) is one example.

 CI as a Complex Communicative System

If we accept the social system theorist Niklas Luhmann’s (1995) argument that social systems are communica-
tive systems, and his useful and compelling definition of complexity, community of inquiry as a communicative 
system may be understood as complex, in that it contains more communicative possibilities than can be actual-
ized. For example, on the level of group interaction there are always more themes than can be explored, more 
questions than can be pursued, and more concepts than the group can take up for inquiry.  On the individual 
level, there are always more possible interventions than one can actually make.  Community of inquiry shares 
many important characteristics that Cilliers believes are common to all complex systems : rich, interconnected 
and nonlinear interactions; recursion, redundancy, feedback, adaptability, and autopoiesis. (1998). I provide a 
brief description of those characteristics below, in the context of the dialogical format of a community of inquiry 
(CI hereafter).

 
Davis and Simmt (2003), in describing the emergence and the functioning of a learning community of 

pre-service teachers as a complex system, emphasized eight characteristics of such a community: internal diver-
sity, redundancy, decentralized control, organized randomness, neighboring interactions, regulatory constraints, 
emergent diversity, and distributed control.  I would diverge from Davis & Simmt’s description of a learning 
community in complexity terms, in particular in the interpretation of two characteristics-- decentralized control 
and neighboring interactions. This is due to the difference in the forms of the communities we describe. Davis 
&Simmt (2003) portray the forming of a learning community of pre-service teachers who have organized them-
selves into study groups, thus forming a network of connected study clusters. None of the teachers were in charge 
of the collective, any more than one individual may be said to be in charge of an ant colony, the growth of a city, 
an immune system, or the functions of the brain. These are all decentralized systems, without a central authority. 
The behavior of any element in the system, and thus the system in general, is influenced by the interactions in 
the systems, and particularly by communication with the most closely located elements, or by the nearest neigh-
bor interactions. Thus, if an ant encounters a high rate of nest-maintenance workers entering and leaving the ant 
colony, there is a higher probability that it will adopt a nest-maintenance job, and if it encounters ant foragers 
busy lugging food home, the chances are it will switch to foraging (Mitchell, 2009). Similarly, in human learning 
communities learners are greatly influenced by communications with other learners from the same small cluster 
(Davis & Simmt, 2003).

In comparison, in a community of inquiry every participant is potentially connected to every other, and since 
all communication has the same range, every communicative intervention is potentially equally influential.  Also 
CI might be better understood as a complex system governed by certain communicative meta-rules from the 
start, where the facilitator is the central communication organizer, the one who enforces those rules. Thus the 
term decentralized control, often used to describe complex systems, does not do justice to the control mecha-
nisms in CI.  The facilitator in a CI exercises control procedurally—she may act to regulate the distribution of turns 
or the length of a speaker’s intervention, call for paraphrasing, summarization or location of the argument, or 
insist on a group response to an intervention that has been over-ridden by the next speaker.

Although it may appear that it is the facilitator who is navigating the direction of the group, unilateral 
navigation is in fact only possible, if at all, in a closed, controlled system, in which the teacher dictates events 
and meanings, and inhibits any dynamic emergence.  Her operative goal is in fact to enable as much as possible 
communicative diversity and clarity, in the interest of acquiring new meanings, and better participation. In fact, 
we could say that procedural control, because it holds everyone responsible for adhering to the communicative 
rules, enables a CI to work towards its regulative ideals-—distributed thinking, distributed intelligence, and distributed 
agency. The term ambiguous control  (Lushyn & Kennedy (2002) indicates that it is neither centralized nor decen-



ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

15

tralized control, but a more complex control mechanism that facilitates the organization of a complex system of 
communication. 

In a collective such as CI, each intervention is understood as potentially important, and it is never exactly 
clear how much and to what extent an intervention might affect the course of the inquiry or its group dynamics.  
The facilitator may have more perceived authority to steer the group, but other members—those less constrained 
by the internalized hierarchy of the traditional model—may exercise their distributed authority from the start. 
This should be welcomed by the facilitator, whose long-term goal is the distribution of power within the system, 
just as she encourages the emergence of distributed thinking and distributed self-expression. In a more mature 
CI, the facilitator also strives to distribute procedural authority throughout the group—that is, to promote a 
system-condition in which all participants, including herself, share in such things as regulating the distribution 
of turns, calling for summarization or restatement, calling for definitions or examples, encouraging alternative 
hypotheses, identifying unstated assumptions, and similar activities.  Neither distributed control nor ambiguous 
control imply that control is equally distributed or that relationships between the members of the group are 
symmetrical; nor does it mean that those relationships are of domination or exploitation, although this can-
not be guaranteed. Non-linearity, asymmetry, power and competition are inevitable to the dynamical nature of 
complexity, and necessary for maintaining it (Cilliers, 1998). If relationships were symmetrical and power was 
equally distributed throughout, the system would quickly lose energy and stagnate. Asymmetry insures shifts in 
the “weight” of contributions, and in levels and distributions of communicative control and power.

Ideally, interactions in a community of inquiry are “rich” and interconnected in the sense that every inter-
action influences and is influenced by others. CI deliberations are nonlinear, in that some interventions are 
amplified, and have a greater effect than others. This is usually described in complexity theory as the “butterfly 
effect,” in that small causes can have large effects and vise versa. Any member of the group can also intervene and 
change the direction of the inquiry with new substantive or procedural suggestions.  The inquiry itself can shift 
direction through a series of individual moves within the “conversation plane,” and the conversation can shift 
“vertically” between inquiry, reflection on the inquiry, and feedback. The interactive dynamics are constantly 
changing—whether in levels of participation or levels of clarity of the “argument.” This presents a picture very dif-
ferent from the traditional monological, unidirectional model of classroom conversation, in which, when there 
is interaction, it is most typically between teacher and individual student.

Complexity thrives on the ambiguous, just-barely-in-control interplay between CI group members, in which 
the pattern of intervention is recursive—new interventions connect with previous ones and may even appear to 
repeat them, but their measure of difference opens possibilities for future interventions, thus forming patterns of 
argument that make of the structure of CI an emergent constellation of interactions.  In fact recursion is a prima-
ry characteristic of this form of collective dialogue, demonstrated through members’ repetition of utterances, or 
through paraphrase and summary; recursion is thus associated with redundancy, which assures through repeti-
tion that what might not have been well-articulated, or which remains ambiguous, can be communicated again.

 
Recursion is a sign of redundancy, which means that there are informational overlaps and alternative com-

municative routes that insure the system’s flexibility.  Because the system has a memory, and each event informs 
consecutive events, the system’s history is not reversible. Recursion and redundancy are carried by the feedback 
loops that are endemic characteristics of complex systems. A negative feedback loop has the function of maintain-
ing the system’s stability through inhibition and stabilization. For example, a negative loop is initiated when 
the facilitator or another member suppresses, on the grounds that it is not relevant to the current inquiry, an 
intervention initiated by a group member that has the potential to open a new direction, or promises to change 
the current group focus. Positive feedback introduces more diversity—for example, when a member suggests a 
new perspective on the issue at hand, thus stimulating reconceptualization or consideration of a new perspec-
tive.  As a whole, the system operates far-from-equilibrium; equilibrium is in fact “enemy” to the system in that 
it represents stagnation or actual system-death.  As such, the system is fed by a flow of new perspectives, new 
information, and new meanings.  Any stability that is achieved is of a dynamic nature, in a communicative zone 



ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

16

that is between chaos (too many disconnected interventions, to the point that meaning-making is impossible) 
and stagnation (no new information at all) (Kennedy & Kennedy, 2010). It is said that complex systems involve 
an interplay between chaos and non-chaos, and that they operate at “the edge” of chaos (Kauffman, 1995; Ba-
ranger, 2000).

All biological and social systems exhibit the two basic characteristics of cooperation and competition. An 
important insight that comes from a new understanding of living systems is the recognition that networks are 
basic patterns of their organization, and that a network’s complexity involves an interplay between coopera-
tion and competition (Capra 1997; Kauffman, 1995). Social systems are networks of organisms; organisms are 
networks of cells, organs and cells are networks of molecules, and they survive because of the intricate coupling 
between cooperation and competition, where often cooperation on one level underlies competition on another, 
or the other way around.  For example, members of a CI are typically competing for the opportunity to shape 
the character of the discussion with their own ideas, but do so in the interest of collaborative reconstruction of 
the concepts in play.

CI as a communicative system is always reproducing itself, in a process that has been termed  “autopoiesis” 
(from “auto” self and “poiesis” production) by the biologist and philosopher Humberto Maturana (Varela, 
Maturana & Uribe, 1974).  Along with the production of utterances that constitute judgments, arguments, 
and sequences of argumentation, there is the autopoiesis of ideas—new meanings that co-evolve with each indi-
vidual’s thoughts, feelings and expectations.  However, all events in the communicative process are discrete and 
temporal. Thus the autopiesis of the communicative system as a whole depends on the ongoing disintegration 
and reproduction of the materials and the catalysts necessary for the continuation of the process. The autopoi-
etic process constantly increases the complexity in the system, which in order to survive struggles to reduce 
complexity. Out of the necessity for the system to maintain an autopoietic balance in CI, just as among any 
interconnected elements in any living system, there emerges the necessity for regulatory constraints. Constraints 
are limiting, but are also enabling.

For example, selection is a constraint born from the impossibility of connecting every intervention or com-
municative event to every other one in the course of the communicative process. If there is no selection, then the 
proliferation of communicative moves and events will amount to chaos in the system.  Selection is both enabling 
and constraining for the production of new ideas and meanings, and therefore must steer its way through the com-
plexity of communication. Any order in a communicative event emerges through a reduction of complexity and 
through the selective conditioning of this reduction. Without such selections in the communicative process, the 
CI system would quickly reach incomprehensible complexity.

The emerging communicative structure’s self-organizing process is not controlled or predetermined in any 
way, and is a response to the system’s environment, or context. Ashley (1962) points out that each dynamic 
system generates its own form of self-organization in response to its need for survival and for the management 
of complexity. Only by structuring complexity and imposing constraints can a communication system acquire 
“internal guidance,” and make self-reproduction possible. This is what ensures the quality and connectivity of 
the communicative events. In that sense, CI develops an organization (structure) of communication that is in 
unison with the general CI meta-rules that ensure balancing of weight, distribution and good communication 
“traffic” of all communication moves, but yet is always unique to the particular system.

Complexity is fed by the diversity in the system, but as Page (2011) notes, it’s not so much the fundamental 
diversity or the diversity of the members of the group that is important for complexity.  Rather, what is essential 
is the emergent diversity in the system, which in CI is manifested at higher levels—that is, in the communicative 
products of ideas, linguistic expressions, restatements, and summarizations, the recombination of which pro-
duces complexity.

Because the elements of the system are interconnected and adaptable, the system as a whole is self-balancing, 



ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

17

and although there may be crises and breakdowns, the latter represent—potentially anyway—transitions to a new 
state of dynamic balance. The autopoietic function finds expression in the idiosyncrasies of the group-as-a-whole 
and the interactions among its members. Notwithstanding its systemic contingency, for any outcome could al-
ways be otherwise, CI as a communicative system is quite robust, and its emergent structure relatively stable in 
time and across contexts.

Acknowledgment and respect for complexity
 
The modern (traditional) way of encountering complexity has resulted, we have discovered, in reductionism 

and determinism, and in the attempt for a totality of descriptions, truths, and understandings in an attempt to 
separate, isolate, analyze, predict, and master. Such an approach is equivalent to occupying an epistemological 
position from which the continuous shifting of relationships in a system are ignored and, in fact, complexity is 
avoided. An acknowledgement of complexity requires that we take into account our new understanding of com-
plex systems and their main characteristics and behaviors. But “taking into account” such knowledge will require 
more than just “knowing about it.” Dealing with complex systems requires, I suggest “respecting the complexity” 
of CI , not only in the dictionary sense of “regard” and “admiration” for the wondrous properties and behaviors 
of complex systems, but also in the etymological  sense of constantly “turning round” and “looking at” (from the 
Latin (“raspicere”) the system  if we want to understand it better (Partridge, 1966). 

  
What I have tried to provide here is a very general description, which does not pertain to the CI model in a 

scientific sense. Any models or descriptions that we create will be flawed, and in ways that we cannot completely 
determine in order to correct them. On the one hand, each member of the system is ignorant of the behavior 
of the whole system—-she cannot “contain” or comprehend the complexity of the entire system (Cilliers, 1998). 
Nor can any model or description predict or control the behavior of a complex system. This is not to say that 
we don’t need theory or description; in fact, complexity theory helps us to better understand the dynamics of 
CI. It also highlights the importance of uncertainty, contingency, local factors, specific context and timing of 
events. Rather than formulating fixed descriptions or models that portray CI as a fixed and rigid system, we 
might do better to devote ourselves to a process of describing, in the interest of understanding it, a system that 
is constantly unfolding and becoming. We can identify no predetermined outcome states, but some states are 
more possible than others. Such states occupy a space that Stuart Kauffman calls the “adjacent possible”—a space 
of possible other states that are close to the current state of the system.  Kauffman describes it this way: “The 
Adjacent Possible is like a forever expanding house, where passing via a particular door from a room to another 
room, opens new doors in the Adjacent Possible which we explore” (Kauffman, 2010, para. 14). 

We cannot grasp the essence of the system in some determinate way, since each description provides a lim-
ited view, and portrays some aspect of the system from a specific position inside or outside it, and at a specific 
point in time. In this sense, the description of the system is also distributed. However, we are bound to “look 
back or again at,” “regard,” “consider” (respicere, respect) the system constantly, appreciating its complexity; to 
take different “snapshots” of the system and attempt to make sense of them, acknowledging the fact that they 
are temporal, local, limited, and unfinished representations, but confident that they can help us to develop an 
ever-fuller understanding of the phenomenon called community of inquiry.

  
Notes

1. The paper was presented at the 2012 conference of the North American Association for Community of 
Inquiry in Vancouver, Canada.

 
References
Ashby, W. R. (1962). Principles of the self-organizing system, in H. Von Foerster and G. W. Zopf, Jr. (Eds.). 

Principles of Self-Organization: Transactions of the University of Illinois Symposium, Pergamon Press: London, UK, 
pp. 255-278.



ANALYTIC TEACHING  AND PHILOSOPHICAL PRAXIS   Vol. 33 Issue 1

18

Baranger, M. (2000). Chaos, complexity, and entropy: A physics talk for non-physicists, http://www.necsi.edu/projects/
baranger/cce.pdf.

Capra, F. (1997). The web of life. Anchor Books. 
Cilliers, P. (1998). Complexity and postmodernism. London: Routledge.
Davis, B., & Simmt, E. (2003). Understanding learning systems: Mathematics education and complexity science. 

Journal for Research in Mathematics Education, 34 (2), 137-167.
Habermas, J. (1990). Moral consciousness and communicative action. Cambridge, MA: MIT Press.
Kennedy, N. & D. Kennedy (2010). Between chaos and entropy:  Community of inquiry from a systems perspec-

tive. Complicity:  An International Journal of Complexity and Education, 7, 2. http://ejournals.library.ualberta.
ca/index.php/complicity/index

Kauffman, S. (1995).  At home in the universe-The search for laws of self-organization and complexity. Oxford University 
Press.

Kauffman, S. (2010). Re-imagining society: Are we trapped by old ideas?   Retrieved from http://www.npr.org/
 blogs/13.7/2010/11/01/130976163/
Luhmann, N. (1995). Social systems. Palo Alto, CA: Stanford University Press.
Lushyn, P. & Kennedy, D. (2002). Power, manipulation, and control in a community of inquiry.  Analytic Teaching 

23, 2: 103-110.
Mitchell, M. (2009). Complexity: A guided tour. New York: Oxford University Press.
Page, S. (2011). Diversity and complexity. Princeton University Press. 
Partridge. E. (1966) A short ethymological dictionary of modern English, 4th edition.  NY: Macmillan Publishing. 
Varela, F., Maturana, U., & Uribe, R. (1974). Autopoiesis: The organization of living systems, its characterization and 

a model. Biosystems, 5, 187-196.

 Address Correspondences to:
 Nadia Stoyanova Kennedy
 Stony Brook University, NY
 nadiakennedy@verizon.net

 


